ELECTBICITY, MAGNETISM, AND GRAVITATION. 159 



removing the book, deflection again commenced, and the 

 needle continued to perform vibrations of about 10° so long 

 as the sun's rays continued upon it. 



At 12, (noon,) placed the book again so as to inter- 

 cept the solar rays, the needle still vibrating; this gra- 

 dually diminished until it became perfectly stationary, and 

 remained so up to 12h. 7im, Removed the book, and in 

 half a minute the galvanometer vibrated to 5°, and became 

 stationary at 2,^°. Replaced the book, and the needle 

 returned to 1^°. The sun had now so changed his position 

 that the other or north extremity of the helix had also 

 become illuminated. On placing a small book to shade the 

 latter from the sun's rays, the galvanometer needle slowly 

 travelled as far as 0^°, then receded to 5"^ below zero, and 

 kept up a continuous vibration from this point towards zero, — 

 the instrument having been previously adjusted. On inter- 

 cepting the sun's rays the galvanometer declined to zero, and 

 there remained perfectly stationary for five minutes. 



At 12h. 20m., the galvanometer being still stationary, I re- 

 moved the book, and in one minute it had progressed to 2^°. 

 All action now ceased, and the needle became stationary at 

 this point. 



December 6th, 1850. — Observed the galvanometer at 2^° 

 at 12, (noon.) Adjusted it; but observed that whenever 

 the sun began to shine brightly, it became several degrees 

 deflected. At Ih. 20m., on paying more particular attention 

 to this circumstance, T found that by interposing a piece of 

 paper or other screen between the instrument and the solar 

 rays, the deviation was corrected. 



I found that by rectifying the instrument in the sun's rays 

 and then interposing a screen, the needle deviated to the 

 left hand 2i°, — and on removing the screen, it speedily 

 returned to its original station. This experiment I repeated 

 about every half minute for a dozen times, and the same 

 results invariably ensued. The needle always answered 



